Field
Aspects of the present invention generally relate to an image processing apparatus and an image processing method for analyzing presence or absence of abnormalities in a printer and estimating a faulty part, and a program for executing image processing.
Description of the Related Art
In recent years, with the improvement in performance of electro-photographic apparatuses, apparatuses (i.e., image processing apparatuses, such as printers) that realize image quality equivalent to that of printing machines have appeared. When a printer is to be operated in a same manner as a printing machine, it is necessary to maintain high level of image quality. However, the printer may be degraded if the printer is used for a long time under stressful operational conditions, and thus an unusual image (abnormal image) may be output therefrom. A streak generated in a sub-scanning direction or a main scanning direction of the printer can be given as an example of a problem arising at high frequency.
Because such an abnormal image arising due to deterioration and the like is difficult to be detected automatically using sensors, it is often the case that the servicing operation is performed after a user has pointed out such abnormalities. However, it is difficult for the user to verbally explain features of the abnormal image. Therefore, for example, even if a service engineer is informed about a streak generated in an output image, cause of the streak cannot be identified unless the service engineer is provided with detailed information, such as a color, a direction, a position, and a width of the generated streak. Therefore, when the user points out that the abnormal image is output, the service engineer has to visit the installation site to check how the output abnormal image look like. Then, the service engineer estimates a faulty portion by checking the abnormal image, identifies related service parts, returns to the service base once to obtain the service parts, and visits the installation site again to perform a servicing operation. If the servicing operation is performed in such a manner, this can cause not only transportation cost of the service engineer to increase, but also downtime to occur because the apparatus cannot be used until the servicing operation is completed, and thus productivity of the user will be lowered remarkably.
Therefore, Japanese Patent Application Laid-Open No. 2011-137895 discusses a technique in which whether a printer is in failure is diagnosed by determining presence or absence of streaks (white spots) through threshold processing executed on divided regions of a scan image of an image output from the printer.
However, the streaks with low density as well as high density can be generated in the image output from the printer. In addition, the streaks are generated in various widths (thicknesses) depending on the cause of failure. Further, in-plane deviation which cannot be identified visually, which is specific to an electro-photographic printer, is generated.
Further, similar to the case of the printer, a scanner has a characteristic in which specific in-plane deviation can be generated in a scan image acquired from read data. Furthermore, the scanner also has such a characteristic that variation caused by the influence of noise occurs in a scan image even if uniform data is read thereby.
As described above, even if the image appears to be uniform when seen by the human eyes, signal values of the scan image vary considerably because of the characteristic of the scanner, and thus it is difficult to detect streaks having different widths or densities using the threshold value.
Because of the above-described reasons, with the conventional technique, it is difficult to determine presence or absence of streaks in an image output by the electro-photographic printer using a scanner.